Dao Companion to the Philosophy of Xunzi

Transcription

1 Eric L. Hutton Editor Dao Companion to the Philosophy of Xunzi

2 Editor Eric L. Hutton Department of Philosophy University of Utah Salt Lake City, UT, USA Dao Companions to Chinese Philosophy ISBN ISBN (ebook) DOI / Library of Congress Control Number: Springer Science+Business Media B.V This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Printed on acid-free paper This Springer imprint is published by Springer Nature The registered company is Springer Science+Business Media B.V. Dordrecht

3 Abbreviations In order to facilitate consistency of references across chapters and to aid readers in locating passages from the Xunzi, the following abbreviations for references are observed throughout this volume. For the Chinese text of the Xunzi : HKCS Lau, D.C., and F.C. Chen, eds A Concordance to the Xunzi. Hong Kong: The Commercial Press. Cited according to the numbering system used in the concordance: chapter number/page number/line number(s). ( Note: Not all authors in this volume follow the exact edition of the text given in this concordance, so the listing of these numbers should not be taken as an endorsement of that edition on their part but is rather primarily for reference purposes.) For English translations of the Xunzi : H Hutton, Eric Xunzi: The Complete Text. Princeton: Princeton University Press. Cited as: page number, or page number.line number. K Knoblock, John Xunzi: A Translation and Study of the Complete Works, 3 volumes (vol. 1: 1988, vol. 2: 1990, vol. 3: 1994). Stanford: Stanford University Press. Cited as: volume number in Roman numerals. page number, sometimes followed by further reference given as chapter number.paragraph number per Knoblock s translation. W Watson, Burton Xunzi: Basic Writings. New York: Columbia University Press. Cited as: page number. 1 1 Note: the pagination of this edition differs slightly from the earlier 1963 edition of Watson s translation. xi

4 Chapter 1 Style and Poetic Diction in the Xunzi Martin Kern The Xunzi is widely recognized as a book of well-developed expository prose, 1 even though its literary style has been called, perhaps unfairly, at best... indifferent (Lau 1970 : 8). Unlike other texts of early Chinese philosophy Lunyu, Mengzi, Zhuangzi, Mozi, and others more it does not pervasively use anecdotes or dialogues to stage or create its purported or real author as a particular persona ( Lewis 1999 : 58); its principal form is that of the discursive essay. 2 Instead of appearing as a speaker in thirdperson anecdotes the way Kongzi, Mozi, or Mengzi do (and have their personas created through these anecdotes), the Xunzi s expository prose speaks from the perspective of arguing for his ideas and against his opponents ( Denecke 2010 : 180). 3 In the third century BCE, the Xunzi thus participated in, and contributed to, the rise of the expository essay that can also be observed elsewhere, e.g., in parts of the Zhuangzi. 4 One must be careful, however, not to overemphasize the text s authorial voice as a 1 Paul Rakita Goldin expresses a common attitude toward the text: Xunzi s writing is succinct and lucid, his philosophical positions original and reasoned (Goldin 1999 : xiii). 2 The only chapters that include mention of Xunzi (in the form of S UN Qing, S UN Qingzi, or X UN Qingzi ) are 8 ( Ru xiao ), 15 ( Yi bing ), and 16 ( Qiang guo ). In addition, the Eulogy (Knoblock ) on Xunzi that may have come from a disciple or later scholar and is appended to the final chapter 32 ( Yao wen ) mentions him as S UN Qing. Because they speak of X UN Kuang in the third person, these chapters are often taken as the works of his disciples. 3 See also William G. Boltz: [L]iterary or essay-like texts, authored by a single writer, in the way we typically think of a text in the modern world, do not reflect the norm for early China but were, at best, the exception (Boltz 2005 : 59). 4 Denecke might be overstating the case for the Xunzi when noting that its new rhetorical format, the expository essay, constituted a radical change, an innovation that was to fundamentally alter the face of Masters Literature (Denecke 2010 : 180). M. Kern (*) Princeton University, Princeton, NJ, USA mkern@princeton.edu Springer Science+Business Media B.V E.L. Hutton (ed.), Dao Companion to the Philosophy of Xunzi, Dao Companions to Chinese Philosophy 7, DOI / _1 1

5 2 M. Kern personal one; 5 in many instances, the seemingly first- person pronoun is not a firstperson pronoun at all but a general one that should be understood as you or one (Harbsmeier 1997 : ). Either way, the expository chapters of the Xunzi reflect a discursive and sometimes even combative style of argument that straightforwardly addresses a series of topics and exceptional in early Chinese rhetoric does so with mundane pugnaciousness (Harbsmeier 2001 : 883). The individual chapters of the received text arranged first by LIU Xiang (79 8 BCE) and then by Y ANG Liang (ninth century) appear as separate monographs on a range of issues, even though they are rarely, if ever, coherent from beginning to end. 6 By examining in some detail a certain number of representative chapters, the present essay argues for a more nuanced appreciation of the Xunzi s style; specifically, by tracing the considerable stylistic differences between individual chapters, it calls the idea of a single homogeneous Xunzi style into question. From the perspective of style, the text emerges more as an anthology of varied writings of Xunzian thought if we admit to an overall philosophical coherence of the text than as a unified work. Thus, we may either allow that XUN Kuang employed a considerable range of distinctive styles in his writings, or we may need to reconsider and broaden our ideas about the authorship of the Xunzi (or both). While the present essay is not the place to address questions of authorship and authenticity, it still offers observations that might be useful to any such discussion. To begin with, the core of the Xunzi is considered to comprise chapters 1 through 24; by contrast, the final eight chapters seem considerably different in nature. As scholars attribute the first twenty-four chapters to XUN Kuang and implicitly assume their authorial unity and coherence, the later chapters have often engendered serious doubt. 7 The chapters in question include two separate sets of poetry (chapters 25 and 26, Cheng xiang and Fu ); 8 one chapter of (in Knoblock s count) 115 brief maxims (chapter 27, Da lüe ), four chapters of Kongzi lore in the style of the Lunyu (chapters 28 31, You zuo, Zi dao, Fa xing, and Ai gong ), yet with just a single brief parallel in the received Lunyu ; 9 5 Here, I disagree with Denecke s analysis as well as with Knoblock s translation. 6 In his introduction, Knoblock offers an extensive discussion of the textual history of the text (K I ). 7 In this respect, the Xunzi is not different from many, if not most, other texts of the early philosophical tradition. 8 For a detailed analysis of the fu poems see Knechtges ( 1989 : 1 31); for a brief discussion of both the working songs (cheng xiang ) and the fu, see Denecke ( 2010 : ); for the Cheng xiang chapter alone, see Malmquist ( 1973b : 63 91) and Malmquist ( 1973a : ). Scholars disagree as to whether the content of these chapters is philosophical (Knoblock) or not (Knechtges). For discussions of rhyme in the Cheng xiang chapter, see Li ( 2010 : 89 93); Zhu ( 1957 : 42 47). 9 That parallel is the brief phrase zhi zhi yue zhi zhi (Knoblock 29.6) that appears as zhi zhi wei zhi zhi in Lunyu 2.17 ( Wei zheng ). By contrast, the four Xunzi chapters of Kongzi lore have numerous parallels especially in Hanshi waizhuan and Kongzi jiayu, and to some lesser degree in Da Dai Liji and Shuiyuan.

6 1 Style and Poetic Diction in the Xunzi 3 and a final chapter 32 ( Yao wen ) that contains anecdotal lore regarding both Kongzi and other early culture heroes. These eight chapters are extremely diverse, with the two poetry chapters in both form and content showing clear affinities to the southern fu style associated with the Chuci ( Lyrics from Chu ) while also being related to Warring States and early Han shui ( attempts at persuasion ) (Knechtges 1989 : 21 31). It may well be for this reason and especially for the topos of the frustrated man bu yu ( not meeting his time ) when the world is morally corrupt and in a perverted state ( Knechtges 1989 : 21 31) that in the bibliographic chapter of the Hanshu, Q U Yuan (trad BCE) and Xunzi are presented as the two originators of the fu genre. According to the account preserved in the Hanshu, the genre at once originated with and culminated in the works of these two authors, descending into a quick decline immediately thereafter (Hanshu 1987 : , 1756). 10 Despite its title, the present analysis will focus on the first twenty-four chapters in the Xunzi, leaving the Cheng xiang and Fu chapters aside together with those that follow them. There are several reasons for this decision. To begin with, the heterogeneous nature and multiplicity of styles in these chapters has long been acknowledged. One would be hard pressed to argue that the Cheng xiang and Fu chapters belong to the core of the Xunzi. It was for sound reasons that LIU Xiang had relegated them to the end of his Xunzi compilation (K I ), and even YANG Liang, who called them za ( miscellaneous ), placed them in the uneasy position between what he considered the authentic writings by XUN Kuang and the additional body of material (chapters 27 32) that he attributed to later disciples (K I.112). 11 Compared to the discursive Xunzi of chapters 1 24, the two poetry chapters seem curiously out of place, and their designations as Cheng xiang and Fu are dubious: while Cheng xiang is simply taken from the first line of the chapter and obscure in its meaning (K III.169), 12 the designation Fu did not originate with XUN Kuang but was quite possibly chosen by LIU Xiang (Knechtges 1989 : 14 15).13 Moreover, it appears that the two poetry chapters were not part of the Xunzi before LIU Xiang, or they were considered entirely marginal: unlike the majority of Xunzi chapters, they do not have any parallels in Western Han literature save for a single snippet from the Fu chapter that is quoted in the Zhanguoce (K I.105) The Hanshu bibliographic monograph Yiwenzhi, which in abbreviated form represents the catalogue of the imperial library at the end of the first century BCE, attributes twenty pieces of fu to Xunzi. 11 The arrangement of presumed inauthentic material at the end of a Masters text is, of course, standard procedure and hence expresses unambiguously what both L IU Xiang and YANG Liang thought of the closing chapters. 12 For a more contextualized discussion of the term cheng xiang, see Kern ( 2003 : 407 9). 13 Simply put, there was no literary genre called fu in X UN Kuang s time; see Kern ( 2003 : ). 14 The quote is in the chapter Chu ce si, Ke shui Chunshen jun.

7 4 M. Kern The more interesting and more important reason to focus on chapters 1 24, however, is a different one: poetry in the Xunzi is not simply what can be found in chapters 25 and 26. Just as in many texts of expository zhuzi ( masters ) prose from the Warring States, it is not a certain body of text bearing the stylistic distinctions of a particular genre; instead, it is a mode of speech, or language use, that deeply pervades what is usually taken as expository prose. 15 This mode of speech is ubiquitous in the Xunzi. To some extent, though not nearly sufficiently, Knoblock s translation shows efforts to identify those poetic passages that are distinguished by rhyme and meter. While still inadequate, this effort marks an important step forward, considering that scholarship on early Chinese philosophy and intellectual history has only recently begun to attend to linguistic form as important for thought and argument. 16 The occasional and sometimes even pervasive poetic style in the Xunzi, and in so many other early Chinese texts of expository prose, is neither a genre nor some sort of superficial, external embellishment of reasoned discourse (let alone impediment to logical expression and interpretation, or, as Angus C. Graham has noted for the Zhuangzi, a collision of logic and poetry ) ( Graham 1991a : 214). Instead, this style of diction is also an intellectual style. It is the very medium through which large parts of the argument operate in the Xunzi, and as such it fulfills as style in any rhetorical tradition functions of persuasion and even of what philosophers are wont to call illocutionary force. Simply put, for the Xunzi and other early Chinese texts, to speak in verse is to speak in the voice of traditional authority and of an emphatic claim for truth. Precisely because the Xunzi is considered a text driven by the desire for arguing, and because that arguing is not simply logical or analytic, its literary style is central to the quality not merely of its form of expression but of its argument itself. Linguistic rhythm itself, as the youshui ( persuaders ) of the Warring States period knew very well, carries a stylistic type of persuasive power by its mere formal structures of parallelism and repetition; this is especially true for a style of 15 Günther Debon has repeatedly pointed to the presence and significance of rhymed sayings ( Spruchdichtung ) in early Chinese expository prose; see Debon ( 1996 : 36 42); on rhymed sayings especially in the Xunzi, see Debon ( 2002 vol.1: 21 30). 16 In discussing the fallacies of the rambling mode in translations of the Zhuangzi, A. C. Graham has offered the most cogent critique of negligence toward the poetic features of early expository prose, summarized in the sentence The effect of assimilating the verse to prose is almost always catastrophic ; see Graham ( 1991b : , esp ). The groundbreaking work on rhyme in early expository prose is Jiang ( 1993 ); see also Long ( , repr. 2009: ), and Tan ( 1995 : 12 19). The gradually increasing body of scholarship on the formal aspects of early Chinese philosophical texts includes Rudolf G. Wagner s analysis of Interlocking Parallel Style, see Wagner (2000 : ), Dirk Meyer ( 2011 ), Raphals ( 1994 : ), Roth ( 1999 ), Queen ( 2008 : ), Baxter ( 1998 : ), Fischer ( 2009 : 1 34), Boltz ( 2005 : 50 78), Liu ( 1994 ), LaFargue ( 1994 ); see also Kern ( 2014 ), Xu (1990 : 58 64), and Morrison ( 1981 : ). Aside from Knoblock s translation of the Xunzi, Hutton s new Xunzi translation also marks off the poetic parts of the text, as does the translation of the Huainanzi by John S. Major et al. (Major et al ).

8 1 Style and Poetic Diction in the Xunzi 5 arguing, ubiquitous in the Xunzi, that is built around analogical patterns and in this conveys a strong sense of order. 17 It is therefore that the present essay focuses on the use of the poetic style in the discursive core chapters of the Xunzi. Of course, concerns about the Xunzi s style go much beyond poetry as a particular type of language use. What the text lacks in literary flourish (especially by comparison with the Zhuangzi ) or historical anecdotes (when compared to many other early philosophical texts), it often gains in focus and stringency of argument, as the individual chapters do tend to focus on their respective subject matter at least for their larger parts. This relative stringency, combined with an explicit diction that rarely uses esoteric anecdotes and elliptic sayings, makes the Xunzi into a text that is relatively easy to follow; it grants few of the pleasures of reading the Zhuangzi but also provokes none of the frustrations the Lunyu stirs in readers hoping to decipher the meaning of some particular passages (not to mention their position within a larger philosophical context). While scholars have considerable difficulty in situating the Xunzi s philosophy between Confucian and Legalist, occasionally resorting to phrases such as realistic or authoritarian Confucian ( Rickett 1985 : 3, 249, 412), they do not face the sort of wide-ranging diversity of thought that forces them to assign any of the twenty-four core chapters to different authors or schools of thought. In short, the first twenty-four chapters of the Xunzi are commonly taken as mutually supportive and non-contradictory, expressing different aspects of a single coherent system of thought; their sometimes divergent viewpoints have been rationalized as coming from distinct periods (early, middle, late) of XUN Kuang s long life. 18 On the whole, it also appears that while the text draws on a considerable amount of traditional source material, including numerous pieces of proverbial wisdom and rhetoric (K I ), there is no direct evidence that it is pieced together from materials common to a wider range of texts. In Knoblock s view, the relatively large number of sections that the Xunzi shares with the Hanshi waizhuan, the Da Dai Liji, and the Liji is by and large the result of these texts borrowing from the pre-liu Xiang Xunzi material, and not the other way around (K I.105 6). By implication, the Xunzi is then generally seen as (a) having existed in more or less its present form before the early Han and (b) being not a composite or compiled text but a truly authored and original one. 19 These widely shared assumptions do not imply that the 17 Here, I allude to Ernst H. Gombrich for the power of formal structure in argument (Gombrich 1979 ); see also Bagley ( 1993 : 34 55). For early Chinese rhetoric, see further Schaberg ( 2001 : 21 56). As Schaberg observes, the rhetoric of good order applies to both speech and written prose; I would add that expository prose with its implied authorial voice is indeed built upon the techniques of persuasion that are first visible in discursive speech. 18 See Knoblock s appendices Composition of Each Book in each of his three volumes. 19 On composite texts as a common phenonemon during the time of the Xunzi, see Boltz ( 2005 : 50 78). In saying truly authored and original, I do not overlook that many scholars (e.g., Knoblock ) have noticed what they call borrowed elements in the text. But to consciously borrow existing language is an intense form of authorship as it implies thoughtful and intentional decisions on the side of the author.

9 6 M. Kern book was initially devised as a grand, unified whole as was the case with part of the Lüshi chunqiu (dated in part to 239 BCE) and the entire Huainanzi (139 BCE); there is evidence that at least some Xunzi chapters, or individual paragraphs, existed independently from their present context in the book. In fact, we have no reason to assume that XUN Kuang thought of his writings as constituting a book. Any attempt to see a particular order in the arrangement of the existing chapters is defeated by the fact that the received Xunzi represents YANG Liang s re-arrangement of LIU Xiang s earlier compilation, which in turn was not the original Xunzi designed by XUN Kuang himself a thing that most likely never existed in the first place. By necessity, the object of our analysis is the received text, with at least some of its chapters being internally in disarray. It may well be that some of the stylistic incoherence is the result of textual deterioration at an early stage, perhaps comparable to what happened, say, to the Ziyi ( Black Robes ) text where the received Liji version is decidedly inferior to the two manuscript versions from Guodian and in the Shanghai Museum corpus, both dating from around 300 BCE. 20 Be this as it may, it remains significant to observe distinct differences of style between and within the individual chapters of the text. In some brief but illuminating comments on chapter 1, Knoblock has argued that the first seven sections of the chapter (in his numbering, equaling roughly half of the chapter) are replete with traditional material that is widely attested in other works dating from Xunzi s time and later while sections eight through fourteen are mostly the original composition of Xunzi and as such [are] much more rarely quoted (K I.124). 21 Such a conclusion should be phrased more carefully: whether or not the second half of the chapter is indeed the original composition of Xunzi is, in fact, impossible to decide. What cannot be disputed, on the other hand, is the fact that by and large, the latter half of the chapter is not shared with other texts, while the first half overwhelmingly is. Why? To some extent, the answer to this question may be found in the analysis of style and such an analysis further reveals that the two halves of the chapter have little in common and perhaps should not be conceived of as an integrated whole. In the present essay, I examine chapter 1 in some detail. This chapter shows significant stylistic features one also finds elsewhere in the Xunzi. Following this analysis, I comment briefly on specific features in several other chapters that are generally considered of central importance to the text as a whole. Whenever I quote from the original, I arrange the text in a way that reveals its formal structures. 20 See Kern (2005: ), and Kalinowski ( : ). 21 In the present essay, I do not always follow the divisions into sections as given in the CHANT version (which is also largely coherent with Knoblock s division). I indicate where I differ from CHANT or Knoblock.

10 1 Style and Poetic Diction in the Xunzi 7 HKCS 1/1/3 5: : A junzi says: In learning, one must not desist., ;, Blue is taken from the indigo plant, yet it is bluer than indigo. Ice is made from water, yet it is colder than water.,,,,, A piece of wood as straight as a chalk line can be rounded [by steaming] to become a wheel; its curvature [will then] conform to the compass. Even when dried in the sun, it will not return to its [former] straightness. The process of rounding by steam has caused it to be like that., Thus [it is said]: If wood is aligned to the chalk line, it will be straight; if metal is put to the whetstone, it will be sharp., If the junzi studies broadly and daily inspects himself on three counts, 22 his understanding will be clear and his conduct without transgression. 22 My translation follows YANG Liang s commentary and the parallel in Lunyu 1.4 ( Xue er ): Zengzi said: I inspect myself daily on three counts ( : ); later commentators on the Xunzi have interpreted the word can (*N-sʕrum) not as san (*srum; on three counts ) but as yan (*m-qʰr[a]m-s; to examine ) and have further argued that the two characters xing hu ( inspect plus directional preposition at ) are an interpolation. Thus, Knoblock translates as the gentleman each day examines himself (135). I see no need for this emendation, nor can I think of a good explanation for the purported interchangeability of can and yan.

11 8 M. Kern The passage begins with the quotation of what David Schaberg has called a platitude persona (Schaberg 2005 : ), namely, the figure of the anonymous and unspecified junzi that appears also in numerous other texts of the time and, as Christoph Harbsmeier has noted, in general does not refer to any specific individual. 23 The quotation of his saying In learning, one must not desist is a gesture toward tradition: whoever the author of the chapter is, his opening words are not in his own voice but draw on pre-existing authority that, furthermore, is not located in a particular person but in a generalized junzi. His statement of received learned opinion is then followed by two sets of analogies: the first, on blue/indigo and ice/water is taken from the natural world; the second, and much more extensive one, is from the realm of craftsmanship that also figures prominently in the rhetoric of other early philosophical prose (De Reu 2010 ; Major 2014 ). Following the second of these analogies, the text returns to a general, indeed apodictic, statement on the matter of learning : If the junzi studies broadly and daily inspects himself on three counts, his understanding will be clear and his conduct without transgression. In this sequence, the analogies in the middle part lead from the initial piece of traditional wisdom toward a statement on learning as transformation of the self by regular exercise of self-examination. The middle part is not built on explicit deductive logic but rather on the implicit inference from analogies, reinforced by repetition and parallelism, that by mere accumulation generate some rhetorical force. The first analogies of blue/indigo and ice/water, for example, are ready-for-use, disposable items from the general store of rhetorical analogies; the second analogy the wood bent by steam and then remaining bent even when dried again is a more original comparison to a person s permanent transformation by learning. It is followed by gu, an introductory sentence adverbial that often does not have a strong logical force (as in therefore ), as it does not function as the hinge between the immediately preceding sentence or section and the subsequent one ( Gassmann and Behr 2005 vol.1: 96). Instead, it frequently serves as the introduction of another piece of traditional wisdom: what follows gu (which I translate as thus [it is said] to indicate that the following is again a quotation or otherwise marked speech) 24 is a general maxim, usually bound by rhyme or rhythm, that is supported by the preceding illustration. Here it is important to remember that we are not in the style of deductive reasoning: while the maxim (in this case not rhymed, but a formulaic couplet governed strictly by parallelismus membrorum ) picks up the analogy of wood, it actually takes it into the opposite direction. Now, wood is not bent but straightened, because this is how it is parallel to the knife that is sharpened. Finally, the text returns to the junzi but not necessarily to the one who was quoted in the beginning, and his theme is not as in the initial proverb-style saying learning that shall never end; instead, the focus is now on the regular practice of selfexamination by which the junzi will permanently transform himself. However, for 23 Personal communication, January On different types of quotation (and pertinent references), see Kern and Hunter ( forthcoming ).

12 1 Style and Poetic Diction in the Xunzi 9 the analogy to wood to operate properly, one would have to assume that the junzi acts on his inner self as on an object distinct from the examining mind: wood neither examines itself nor bends or straightens itself; it is acted upon so as to become permanently changed. In short, this sequence alone includes five sections that are all (a) mutually independent and (b) formed in different ways. Any of these sections could easily be integrated into different contexts, and one might well want to ask how many of them are original to the Xunzi. 25 What holds the section together is the fact that it is framed by two maxims associated with the junzi and his learning, even though these maxims emphasize different aspects of his self-cultivation. In the conventional division of the text, 26 the first section ends right here. This would be fine if it would not mean that the next section then had to start with another gu the sentence adverbial to introduce a concluding commonplace. Yet the reasoning for starting a new section here is not implausible: the following lines seem, at best, only loosely related to the preceding text. Thus, the section introduced by gu may indeed not belong to the first section but in this case, it may be altogether misplaced, or whatever may have preceded it originally is now lost. On the other hand, one might argue that its weak connection to the preceding text in the current version of the Xunzi is symptomatic of the entire first section which, as just shown, is altogether loosely integrated and possibly assembled from various bits and pieces. If the first section in its present form is indeed one author s original composition, it does not show him overly concerned with the cogency of his argument or he relies on an audience of insiders capable of generating from his style a stringent line of thought. Be this as it may, what follows the second gu is another piece of traditional wisdom strictly organized by syntactical patterning: HKCS 1/1/7 10:, ;, ;, Thus [it is said]: If one does not climb a high mountain, one does not understand the height of heaven Here, I differ from Knoblock s assessment which is purely based on the comparison of the Xunzi with other transmitted sources. To say that these sources especially the Da Dai Liji and the Hanshi waizhuan seem to be quoting from the Xunzi and not vice versa is not the same as saying that whatever they quote did actually originate with the Xunzi. 26 As reflected in the CHANT edition as well as in Knoblock s translation. 27 Clearly, the metaphor refers to the height of the sky, yet at the same time, the two sentences here invoke the heaven/earth cosmology.

13 10 M. Kern If one does not look down into a deep valley, one does not understand the depth of the earth. If one does not hear the words left by the former kings, one does not understand the greatness of learning.,,, As for the children of Yu, Yue, Yi, and Mo: at birth they make the same sounds; growing up, they differ in their customs. Education causes them to be like that. :,,, (*-ə) (*-ək) (*-əp) (*-ək) (*-ə) (*-ək) An Ode says: Ah, you noble men, do not consider permanent your being at rest and at ease. Be reverent and respectful of your positions, be fond of those who are upright and straight. Exert [your inner] spiritual capacity and adhere to it, to increase your radiant blessings., Among one s spiritual capacities, there is none greater than the way of transformation. Among blessings, there is none more enduring than being without misfortune. Once again, the passage is structured by rhythm and semantic parallelism; 28 the only rhymes are in the quotation of the final stanza from Ode 207, Xiao ming ( Lesser Brightness ). The initial passage following gu ( Thus [it is said] ) has no discernable connection to anything before or after except for its praise of learning from the words of the ancient sages. This, however, is then followed by an analogy that resonates closely with the earlier metaphor of wood that is permanently bent. 28 The absence of rhyme, however, does not mean the absence of poetry; Baxter notes that both rhyme and semantic patterning, especially including semantic parallelism, are used as poetic devices in the Laozi (Baxter 1998 : 237).

14 1 Style and Poetic Diction in the Xunzi 11 In fact, the lines that conclude the earlier analogy of bent wood and the later one of acquired customs are strictly parallel: The process of rounding by steam has caused it to be like that. Education causes them to be like that. If anything, this direct and likely not accidental piece of parallelism suggests that the two parts of text do belong to a single section, even though there is additional material in it. In other words, the strict parallelism signals both unity and, perhaps, the addition of formerly unrelated material to that unity of expression. What follows the second analogy the quotation from the Odes as well as the final statement on one s spiritual capacity and blessings is only partially related to the main theme of cultivation through learning, namely, in its reference to transformation and possibly also to being zhi ( straight ), but not at all in its reference to blessings. In its strict parallelism, the final statement appears once again as some sort of proverb and was possibly independent from the Odes quotation with which it is paired here. In sum, the first section, or sections, do not develop a cogent argument; instead, they embellish the principal thesis on the lasting influence of learning with various pieces of traditional wisdom culled from different sources. Consider now the second (or third) section of the first chapter: HKCS 1/1/12 15:,, I once spent the whole day thinking, but it was not as good as what I learned in an instant. I once stood on my toes gazing into the distance, but it was not as good as what I broadly saw after ascending a place on high.,, ;,, By ascending a place on high and waving, the arm does not gain in length, yet its visibility reaches into the distance; By shouting with the wind, the voice does not gain in strength, yet its audibility becomes more distinct.,, ;,,

15 12 M. Kern To make use of carriage and horses is not to benefit one s feet but to go a thousand li ; To make use of boat and oars is not to gain ability with water but to cross rivers and streams., The junzi is not different by birth; he is good at availing himself of external things. This section consists of three statements, each composed of two parallel halves, and a concluding pronouncement on the junzi. Each such prose couplet has its own meter and rhythm, yet all three are unified in their extreme, mechanistic parallelism; and one leads to the next through the continuous use of a key phrase ( deng gao in the transition from the first to the second statement, and the negative fei from the second to the third). Each statement is an illustration of being good at availing oneself of external things ; all three are then summarized in the statement on the junzi. What we see here, as before, is an accumulation of examples, a brief catalogue of mutually independent units. Looking back at the first two (or three) sections discussed so far, the recurring element is the mention of the junzi : 1. A junzi says: In learning, one must not desist. 2. If the junzi studies broadly and daily inspects himself on three counts, his understanding will be clear and his conduct without transgression. 3. The junzi is not different by birth; he is good at availing himself of external things. Indeed, if there is a discernable theme in the beginning of the Xunzi, together with the emphasis on xue ( learning ), it is the concern with the ideal of the junzi :29 a person whose status is not inherent or inherited, but earned through effort and the ability to act upon himself and to draw on external things. Importantly, this ideal is an attainable choice and hence can be argued for by way of persuasive rhetoric. Strictly speaking, everything beyond the three statements on the junzi is dispensable in the sense that any part of it could be dropped or replaced by something different. In these cumulative sections, the Xunzi does not develop an explicit deductive argument; rather, the text pronounces itself three times on the junzi and then, in seemingly random order, fills its columns with illustrations and formulaic pieces of traditional wisdom. Strikingly, none of these pieces and nothing in the opening passages involves the style of historical anecdote one is accustomed to 29 Goldin sees this as the theme of the entire Xunzi : The overarching preoccupation that binds together the diverse arguments and reflections in the text is the role of the noble man (Goldin 1999 : xi).

16 1 Style and Poetic Diction in the Xunzi 13 read in other early writings of expository prose, nor is there a single historical reference to anyone. The text here is not organized by chains of deductive arguments and conclusions but also not by the logic of narrative; nor is it in any way adjusted to any sort of historical context. This characteristic is true for much of the Xunzi and shared with a text like the Laozi, but not with most other Warring States writings, including most of the recently excavated manuscripts of expository prose. 30 The fourth (or fifth) section of Exhortation to Learning offers yet a different way of traditional discourse, namely, the use of rhyme that is found in many passages of the Xunzi (Debon 2002 ): without any introduction, this section is composed of tetrasyllabic lines and almost entirely rhymed, invoking the formal patterns of the Odes. The passage falls neatly into four sections of four lines each, which are distinguished by particular syntactic structures, further emphasizing the divisions already marked by rhyme change. These brief sections are four variations on a common theme: 31 HKCS 1/2/3 5:, *-ə A *-ə A, *-ək B *-ək B As the categories of things arise, They always have something from which they begin. As honor and disgrace arrive, They always are manifestations of [the person s] virtuous power., *-uŋ x *-ak C, *-iŋ x *-ak C Meat that is rotten brings forth worms, Wood that is withered produces grubs. When neglecting the self by being lazy and indolent, Misfortune and disaster will arise. 30 Still another characteristic, finally, which separates the Lao-tzu from much of early Chinese philosophical discourse, is that it is entirely free of narration, in the sense that its statements are general and not anchored to any particular persons, times, or places. There is no indication of who is speaking, no direct reference to historical events. This contrasts strikingly with typical Confucian discourse (Baxter 1998 : 240). 31 In my representation of the rhyme pattern, the small letter x represents a non-rhyming line. My simplified representation of the rhymes is derived from William H. Baxter and Laurent Sagart, Baxter-Sagart Old Chinese Reconstruction (Version 1.00), online at Accessed January 23, 2012.

17 14 M. Kern, *-o D *-ok D, *-iŋ x *-o D What is strong gives itself as support, What is soft gives itself for bundling. When vileness and depravity reside in a person, Resentment is what he brings upon himself., *-it E, *-aj F, *-it E *-aj F Where firewood is spread out evenly, Fire will seek out the driest. Where the ground is leveled evenly, Water will seek out the dampest., *-eŋ x, *-an x *-aj x As grasses and trees grow together with their kind, As birds and beasts form flocks, Each thing accords to its own category. Taken together, these rhythmic and euphonic sections offer a series of analogies that illustrate the principle of sympathetic resonance in the natural world: because things respond to one another according to their lei ( categories ) of natural disposition, actions have their specific and inevitable consequences. Rhetorically, the four sections contain what seem to be snippets of conventional wisdom. They are persuasive for two reasons: as observations of the natural world and by the force of sheer accumulation that amounts to a veritable catalogue of phenomena of natural resonance. The altogether eighteen tetrasyllabic lines are capped with a concluding statement of six characters that offers, by a process of induction, the abstraction of the principle illustrated: each thing accords to its own category. From here, the text moves closer to its conclusion, beginning with a summarizing shi gu ( and for this reason ) that leads to yet another set of analogies on the same theme before concluding with a three-line pronouncement introduced again by gu ( thus ) that caps the entire fourth (or fifth) section of the chapter. At this point, the

18 1 Style and Poetic Diction in the Xunzi 15 text maps the social world onto the natural, claiming that we can choose our actions but then cannot control their natural and therefore inevitable and predictable consequences. The junzi must be cautious in speech and action because his behavior may attract calamity according to the same principle of resonance that governs the preceding analogies from the natural world: HKCS 1/2/5 7: ; ; And for this reason [it is said]: Where the archery target is set out, bows and arrows will arrive; Where the forest woods are flourishing, axes and halberds will arrive; Where trees provide for shade, numerous birds will rest; Where [things turn] sour and acid, gnats accumulate.,, Thus [it is said]: When speaking, one might invite disaster; When acting, one might invite disgrace May the junzi be cautious about where to take his stand! Once again, the statement on the junzi which easily matches up with the three earlier ones listed above provides the closure of the entire section; it is as much a reminder of the chapter s topic proper as it serves as a device to structure the text indeed, a kind of punctuation mark. It is remarkable how the individual sections discussed so far are not only selfcontained but also composed in different styles, ranging from what on purely formal grounds of rhyme and meter could be called poetry to the variety of prose patterns. Thus they could be linked not only to different discourses (e.g., about the natural world, the realm of craftsmanship, or moral behavior) but also to different rhetorical figures and patterns of speech. It is unlikely that these passages were original compositions by a single author; more plausibly, they were diverse expressions of traditional wisdom and as such readily available to the compiler of Exhortation to Learning.

19 16 M. Kern It is not surprising that these expressions found their way into a range of different texts (K I ). To give just one example, in Han times the statement (or half- statement, as it is only part of a couplet here) I once spent the whole day thinking, but it was not as good as what I learned in an instant is attributed to Kongzi in the Da Dai Liji and to Zisizi in the Shuiyuan. Considering that already in the Xunzi, all these individual statements are not integrated with whatever follows and precedes them, it is not possible to identify their origin; just as we see them used as bits of traditional lore re-appearing in Han texts, they may well have preceded the Xunzi as well. The same should be held for similar passages across many other chapters of the text; what finds itself as quotation or parallel in Han texts may very well not be derived from the Xunzi but could have been material that was traditional or shared already in the third century BCE and entered the Xunzi as such. 32 The traditional idea of XUN Kuang as the principal origin of his text interprets ideological differences within the Xunzi as coming from different periods of XUN Kuang s life; and in a significant number of cases, it requires the assumption that certain passages are misplaced from an earlier to a later stratum or vice versa. Inescapably, this procedure may well be called a classical case of the biographical fallacy : relying to a considerable extent on the text itself, one reconstructs a coherent author whose intellectual biography then, in beautifully circular reasoning, serves as the master tool to stratify his different ideas chronologically. The text-critical observations made so far can be extended to the entire first half of Exhortation to Learning. In each paragraph, metaphors and analogies from either the natural world or the realm of craftsmanship are lined up in series, no historical references are included, and the passages end with a brief statement on the junzi ; in two of three cases, this final statement can then be found, verbatim or unmistakably related, in the Lunyu. In addition, the Odes are quoted twice, each time explicitly ( An Ode says: ) and with six lines, and the entire text is punctuated by gu and shigu, each time gesturing toward established wisdom that appears both conclusive and unquestionable. Aside from the Odes quotations and the implicit gu and shigu gestures toward traditional authority, no other text is explicitly invoked; thus, it is impossible to decide, for example, whether the Xunzi is quoting from an early version of the Lunyu or whether the latter, at some subse- 32 For a convenient survey of such passages, see the appendices Composition of Each Book in each of Knoblock s three volumes. For materials shared between the Xunzi and various pre-han or early Han texts, Knoblock likewise notes that there is no reason to consider the possibility of direct quotation since we are probably dealing with traditional material ancestral to both the Xunzi and these texts (K I.125). However, Knoblock does seem to assume that where such material is present in the Xunzi, it was consciously selected by X UN Kuang and hence was under his authorial control.

20 1 Style and Poetic Diction in the Xunzi 17 quent stage, possibly during the Han, 33 adopted the lines on the junzi from the Xunzi or a third source. It is only with section seven, 34 a little more than half-way into of the first chapter, that Exhortation to Learning adopts the diction of expository prose, starting with a two-part rhetorical question: Where does learning begin? Where does it end? (??) The answer, introduced by a simple yue ( it is said, or I say ), begins once again with a pair of formalized statements: HKCS 1/3/7 8:, ;, In its sequential order: It begins with reciting the classics, it ends with reading out loud the ritual [precepts]. In its meaning: It begins with being a learned man of service, 35 it ends with being a sage. It is possible that yue, as understood by traditional commentators and translated by Knoblock, means I say. However, yue may well mark the above pair of lines as another saying of traditional origin, 36 especially as the text that follows them continues in free prose. What is emphasized by yue is only the paired statement, not the entire section that follows. Such marking of speech is a common rhetorical feature of early Chinese expository prose in which case yue may indicate that the brief maxim is precisely not in the author s original voice but a piece of wisdom he cites approvingly. By contrast, this maxim is followed by a rare instance (in chapter 1) of several sentences in unbound prose: 33 For a possible Western Han compilation date of the Lunyu, see Zhao ( 1961 : 11 24), W. Zhu ( 1986 : 40 52), Makeham ( 1996 : 1 24), Hunter ( 2012 ). 34 In CHANT; Knoblock s section eight. 35 In social terms, shi refers to the lowest aristocratic rank; in the present context, it implies the learned man of low aristocratic status (still above the unranked commoners) who is associated with military or other service. 36 On the rhetorical use of such markers of direct speech, see Kern and Hunter ( forthcoming ).

21 18 M. Kern HKCS 1/3/8 9:,, By truly building up effort for a long time, one enters [into the process of becoming a junzi ];37 learning is something which continues until death and only then stops. Thus, while the sequential order of learning continues to the end of one s life, when it comes to its meaning, it is what must not be abandoned for even an instant. Those who engage in it are humans; those who abandon it are wild beasts. Without transition, this is in turn followed by a rhythmic and rhymed account of three of the classics: HKCS 1/3/9 10 ; ;, *-ə *-ə *-ə Thus [it is said]: The Documents are the essentials for government affairs; The Odes are where fitting tones come to rest; The Rites provide the great distinctions according to [social] rules, they are the guiding principles of classification. This self-contained unit of three rhymes is then elaborated upon as follows, with the learning of both the Music and the Springs and Autumns Annals added: HKCS 1/3/10 12 (*təʔ) (*N-kək),,,, When learning reaches up to the ritual precepts, it stops. This is what is called the pinnacle of the moral way and its virtuous power. 37 The various commentators cannot agree on the meaning of ru ( enter ) here; see Wang ( 2005 vol.1: 26 27).

22 1 Style and Poetic Diction in the Xunzi 19 The reverence and refinement of the Rites, 38 The moderation and harmony of the Music, The breadth of the Odes and the Documents, The subtlety of the Springs and Autumns : Everything between Heaven and Earth culminates in [learning]. Nothing of this adds up to an actual argument; it is more like a parade of platitudes sputtering forth from the discourse machine 39 that reproduces itself in ever new variations on a circumscribed theme, in this case, learning. But this, of course, is the force of its argument: contrary to the celebration of XUN Kuang as an author with an emphatic personal voice, the text, while certainly advancing an intellectual position, here is emptied of any individuality or surprising thought: in a mantra-like style both rhythmic and repetitive, it falls from one rhetorical pattern into the next, generating and regenerating itself in a continuous and inescapable loop of statements that are asserted but not argued. Because the passage is not built as a linear structure, its continuation is not predicated on what comes before and after any of its parts. It can be rejected but never refuted. HKCS 1/3/14 15,,,,,,,, ;,,! As for the learning of the junzi : It enters the ear, Manifests itself in the heart, Extends across the four extremities, 38 Here and elsewhere, I translate the term li in two different ways: when standing alone, as general ritual precepts ; when being part of a list of what are clearly the liu yi ( six arts ), or some of them, as the title of a text (i.e., the Rites ). By the time of Xunzi, this canon of learning was well-established, as is now proven by the manuscripts from Guodian of around 300 BCE. 39 I borrow the term from Owen ( 2001 : ).